Automatic is More Accurate

Determining the bulk specific gravity of fine aggregate is important when designing hot mix asphalt. The bulk specific gravity is used in calculating the voids in the mineral aggregate of an HMA mixture. The current method of determining dry bulk specific gravity (Gsb) — AASHTO T 84 — uses a cone and tamp to determine the saturated surface-dry (SSD) condition of fine aggregate. This method does not work well when determining the SSD condition of angular or rough fine aggregates because they do not readily slump. Therefore, a more accurate and repeatable method of determining Gsb is needed to provide lower variability between operators and to address problems with angular materials. In order to solve this problem, a method that is more automated and less user dependent is needed to determine both Gsb and absorption of fine aggregates.

During the 1970s, the Arizona Department of Transportation tried to develop a prototype for determining SSD using a rotating vertical tube. Warm air was blown through the tube while it rotated. Using the plots of the inlet and outlet temperature and the basic principles of thermodynamics, they determined the SSD region of these plots. The prototype produced encouraging results, however, it had a high variability.

The National Center for Asphalt Technology continued working on Arizona DOT's ideas. Instead of blowing warm air vertically over the sample, NCAT tried blowing the warm air longitudinally in a steel drum while it was rotating on its side. NCAT discovered that the SSD point could be determined more repeatable by monitoring the outgoing relative humidity. There were several problems with this method, including: inconsistent drying, loss of fines, clogging of screens, aggregate sticking to the drum and the prototype was not automated.

Testing of two automated measuring devices — the InstroTek Corelok prototype and the Barnstead/Thermolyne SSDetect eventually produced quantifiable results that were used to plot consistent and repeatable measurements. Both InstroTek and Thermolyne tried the "dry to wet" approach of obtaining Gsb and absorption. InstroTek used a calibrated pycnometer and vacuum pressure and Thermolyne used an infrared signal to determine SSD combined with a vacuum and agitation system to determine apparent specific gravity (Gsa).

InstroTek devised a method using a combination of a calibrated pycnometer and vacuum-sealing device to determine Gsb and absorption. The pycnometer (volumeter) is used to determine the bulk volume of the sample and the vacuum-sealing device is used in determining the apparent specific gravity. The Gsb is the overall volume of an aggregate particle including the volume of the pores that are filled with water. The InstroTek approach requires that sample be placed into a calibrated pycnometer. The volume of the pycnometer is calibrated by filling it completely with water (before each set of 10 samples). To test a sample, the container is halfway filled with water and a 500-gram dry sample is added. The sample is stirred to remove entrapped air. Additional water is added and a lid is then placed on the pycnometer. The remaining air space is then filled with water. This is used to determine the volume of the aggregate by the displacement of the water. This whole process is done within two minutes to reduce the amount of water absorbed into the pores of the aggregate, thus producing the bulk volume of the fine aggregate.

The equipment of the Thermolyne procedure for determining fine aggregate Gsb and absorption consists of two parts, the AVM unit and the SSDetect device. The AVM unit is an automated device for removing entrapped air from a volumetric flask. The unit includes and automated vacuum source and orbital mixer. A 500 ml volumetric flask is partially filled with water. A 500-gram sample of fine aggregate is added to the flask and the flask filled with water to the calibration mark. The flask is then loaded into the AVM. The AVM removes the entrapped air through the application of the orbital mixing action and partial vacuum over approximately a 16-minute period. After the flask is refilled to the calibration, its weight is determined. The AVM sample is used to determine Gsa and the film coefficient. The film coefficient is a calibration factor for the infrared reflectance measurements made with the SSDetect.